rstudio::conf 2018

This RPres/html is available at Github: https://github.com/gsimchoni/rstudio_conf_2018

• Air Quality
• Stock Market Analysis
• DNA Microarrays
• Mapping Bicycle Routes
• Asteroid Density Estimation
• Bayesian Generalized Linear Mixed Effects Modeling of Tuberculosis
• Sparse Cannonical Correlation Analysis for High Throughput Textual Data from Reuters

• Beyonce
• Ballet Dancing Statistics
• Predicting America's Next Top Model Winner
• Survivor Cast Survival Analysis
• Disproving Astrological Theories
• Recurrent Neural Networks for Rap Lyrics Production
• Estimating the Winner in a 1976 Nadia Comaneci vs. 2016 Simone Biles battle
• Beyonce

• Packages are hard
• Who would want to download my package and run it?
• My code isn't at a level where it can be packaged and distributed
• I'm more of a one-time script kinda girl, moment to moment…
• I hate writing documentation and tests
• CRAN would hate me (a.k.a CRANophobia)

• Gold mine
• ~170 multiple-choice questions regarding risk behavior of youth (9th to 12th grade)
• Asked in huge surveys conducted at the national and state level, every 2 years, from 1991 to 2015.
• “During the past 30 days, how many times did you use any form of cocaine, including powder, crack, or freebase?”

But:

• API is too inconvenient and slow for any real research
• The barcharts are ugly
• Can only download the (“Combined”) data as a “ASCII” raw .dat file or as a MS Access (?!) file
• The raw “ASCII” data file, however, has no column labels, and to get those you need to run a script you can download either in SAS or SPSS

So:

• Download the MS Access driver
• Download the MS Access zipped file, unzip
• Install the survey package
• RTFM (80 pages long PDF)
• And to get a simple answer to a simple question: how many of US youth are currently sexually active…

library(survey)

yrbss_data_binary <- svydesign(id = ~PSU, weight = ~weight, strata = ~stratum, data = yrbss_data_binary, nest = TRUE)

svyciprop(~I(qn63 == 1), subset(yrbss_data_binary, year == 2015), na.rm=TRUE)

• Hilary Parker: Writing an R package from scratch | ~5 min
• Karl Broman: R package primer | ~1 hour
• Hadley Wickham: R packages | ~A few days
• Stack Overflow: Stack Overflow | ~Lifetime

library(tidyverse)

tibble(.variable = rep("qn65", 26),
       .location = rep("US", 26),
       .year = rep(seq(1991, 2015, 2), 2),
       .filters = rep(c("sex == 1", "sex == 2"),
                      each = 13)) %>%
  bind_cols(
    pmap_df(., getProportionSingleVariable)) %>%
  ggplot(aes(.year, prop, color = .filters)) +
  geom_line() +
  scale_color_manual(
    labels = c("Female", "Male"),
    values = c("red", "blue")) +
  labs(title = paste("YRBSS: Proportion of Male",
                     "and Female US Youth",
                     "Reporting Use of Condom"),
       x = "Survey Year",
       y = "Proportion Reporting Use of Condom",
       color = "Gender")

library(ggwithimages)

#...

ggplot(kathmandu_hourly_aqi, aes(hour, aqi)) +
  geom_line_with_image(annapurna, sky) +
  labs(title = paste("Air Quality Index in the",
                     "Thamel, Kathmandu, Nepal"),
       subtitle = paste("Measured in PM2.5",
                        "by the US Embassy",
                        "in Kathmandu"),
       y = "Hourly Mean AQI [PM2.5]",
       x = "Hour") +
  ylim(c(50, 200)) +
  theme(text = element_text(family="mono"),
        axis.title.x = element_text(size=12),
        axis.title.y = element_text(size=12))

• Portfolio
• Exercise & Learning
• Fun
• Fame
• People using your code, appreciating it, improving it (PR)
• Sense of community
• Job offers & Other opportunities
• Even improved my own job